I have a few guesses
- Sometimes, complex cards are genuinely worse because they only happen absent a great idea. Like, if you have a unique idea, you make it into a simple card and are done with it. If you don't have a unique idea, you just start tagging on stuff onto a card (and then maybe scrap/re-add/modify) until you sort of have something. This is not always how complex cards are created, but if it is, it may well be justified that they don't do well
- Complex cards have more components and hence more to criticize. Especially relevant if, to whatever degree, you adopt a measure based on how many flaws you can detect rather than the more subjective sense of how excited you are by the card
- When you design a card, understanding it feels like a trivial cost/buy-in compared to the value that the card provides as an idea or if it were/is played with. But if you judge (or even just read other cards) then you have no prior investment into most designs, so trying to get the hang of it can feel like a significant cost. This can create a negative vibe and since everyone secretly decides everything on vibes, that lowers chances significantly
- As a judge, you don't want it to look like anyone can win the contest by force. E.g., say we make a contest and everyone submits a simple design except one person who makes a 5-step traveler line. If you give the win to that person, it sort of looks like you're rewarding the most work rather than the best idea. (This is probably the weakest point though, and in fact high effort cards don't tend to get a lot of upvotes from other people, either.)
It altogether seems like most people primarily like the "oh that's really clean and clever" type cards.
I feel like there is a bias towards cards that are always relevant in any kingdom, as in cards that fit into most "normal" engine strategies. On the other hand, card that need some more niche conditions to be viable, but don't drift too far into the realm of "this card is garbage unless in this 1% of kingdoms where it is all you care about", are often dismissed because of it.
It is very easy to design a card that doesn't work. It is pretty easy to design a card by looking at cards that already exist and simply remixing the things that are proven to work well and then be reasonably confident the result will also work. It is much harder to design a functioning card that actually has the property above.
I'd argue that the effect is stronger in contests with very open prompts. The more specific the requirements, the harder it is to design an adequate card. Maybe people have a tendency to submit cards that just clear the bar for entry instead of withholding their entry when the deadline draws near, even though they themselves would admit that the card could still use some work.